AVS帧间解码算法的实时实现及优化
摘要
近年来,图像视频编码技术和编码标准得到了长足的进步。AVS是我国第一个具有自主知识产权的数字音视频编解码技术标准,该标准在技术和性能上达到国际先进水平,具有压缩率高、差错恢复能力强、适用范围广等优点,是高清晰度数字电视、高清晰度激光视盘机、网络电视、视频通信等重大音视频应用共同采用的基础性标准。为达到理想的数据压缩率,AVS编码器采用了很多新的编码算法,这些算法极大地增加了编码器的计算复杂性,使其目前在PC环境下的实时应用受到很多限制。要实现AVS的实时编解码,必须在不以性能牺牲为代价的前提下对其进行大量的优化工作。
本课题通过对AVS视频编解码标准的研究和学习,设计实现了基于PC的AVS视频解码系统,并通过后期优化,在并未降低性能的前提下对D1分辨率基本上达到了实时解码播放的要求。本文的工作主要集中在以下几个方面:
1.本文首先研究学习了AVS视频标准的发展现状,系统结构,核心技术和比特流结构。在分析学习AVS标准的同时,将AVS标准与MPEG-4/H.264技术做了比较。
2.从解码系统总体框架、系统工作原理和技术路线等方面,对实时AVS解码系统进行了总体设计和程序代码的实现。并且通过对关键问题的分析,逐步实现了图像头和条带解码,宏块解码、帧间预测等几个核心模块。
3.对解码系统编程实现之后,进行后期优化和系统整体性能测试。主要从内存结构、C代码和MMX三个方面进行优化。针对计算量大、复杂度高的运动补偿和去方块滤波模块,用快速算法和简化算法减少计算量,提高运算效率。并使用MMX,SSE2指令集针对计算量大、频繁调用,但是规则性强的运动估计插值和环路滤波模块,进行优化,提高了计算速度和效率。实现了PC机上AVS解码器实时解码。
最后,对全文进行系统、全面的总结,指出了下一步研究和改善的方向,并展望了AVS的良好应用前景。
Multimedia video coding techonogy and corresponding coding standard are developing rapidly these years. AVS is the first digital audio/video encoding and decoding technology standard of independent intellectual property rights in China. This standard has reached the international advanced level of digital audio/video encoding and decoding standard in technology and performance, and has such advantage as high compression rate, strong erroe resilience capacity, highly adaptability and so on. AVS is a common adopted basic standard in the important application of high definition digital television, network television, video communication and so on. In order to gain good coding result, AVS adopts so many new algorithms that the amount of calculation increase drastically. As a result, its applications in PC environment are restricted. In order to realize AVS real-time codec, many optimization works have to be done. But these optimization gains should not at the cost of obviously coding performance drop.
We mainly studied the AVS standards of video coding/decoding,used hardware and software resources of research team,adopted PC platform,designed a AVS video real-time decoding system.After post-optimization,the average resolution sequence is about 20.8 frame/s,basically achieved real-time decoding broadcast with little performance loss.Our work mainly focused on the following:
Firstly, we studied the AVS standards of video,including development situation,system structure,bit stream structure and its core technology.Through the analysis of the AVS standards,we also made a comparison between AVS standards and MPEG-4/H.264 technology.
Secondly,we finished the general design and implementation of an AVS real-time decoding system,mainly focusing on decoding system framework selection of software and hardware platform,system working mechanism and technology route.Through the analysis of key problems,we gradually finished several core modules like image head and straps decoding,macro block decoding,code streaming parsing and cross-frame forecast.
Third,post optimizing and overall system performance testing after decoding system programming.Optimization is performed in memory structure, C code and MMX.Aiming at the movement compensation and loop filtering modules that contain massivecomputation and complex , we use fast or predigest algorithm to cut down the computation amount and adjust the decoding function to enhance the efficiency.And we use MMX to optimize the modules that are frequently called but well formed ,which would bring a higher performance.Achieved an main profile of PC, real-time decoder decoder.
Finally,A comprehensive summary is given about this paper,and point out the direction of furthor research and improvement,and the promising application prospect of AVS.
本课题通过对AVS视频编解码标准的研究和学习,设计实现了基于PC的AVS视频解码系统,并通过后期优化,在并未降低性能的前提下对D1分辨率基本上达到了实时解码播放的要求。本文的工作主要集中在以下几个方面:
1.本文首先研究学习了AVS视频标准的发展现状,系统结构,核心技术和比特流结构。在分析学习AVS标准的同时,将AVS标准与MPEG-4/H.264技术做了比较。
2.从解码系统总体框架、系统工作原理和技术路线等方面,对实时AVS解码系统进行了总体设计和程序代码的实现。并且通过对关键问题的分析,逐步实现了图像头和条带解码,宏块解码、帧间预测等几个核心模块。
3.对解码系统编程实现之后,进行后期优化和系统整体性能测试。主要从内存结构、C代码和MMX三个方面进行优化。针对计算量大、复杂度高的运动补偿和去方块滤波模块,用快速算法和简化算法减少计算量,提高运算效率。并使用MMX,SSE2指令集针对计算量大、频繁调用,但是规则性强的运动估计插值和环路滤波模块,进行优化,提高了计算速度和效率。实现了PC机上AVS解码器实时解码。
最后,对全文进行系统、全面的总结,指出了下一步研究和改善的方向,并展望了AVS的良好应用前景。
Multimedia video coding techonogy and corresponding coding standard are developing rapidly these years. AVS is the first digital audio/video encoding and decoding technology standard of independent intellectual property rights in China. This standard has reached the international advanced level of digital audio/video encoding and decoding standard in technology and performance, and has such advantage as high compression rate, strong erroe resilience capacity, highly adaptability and so on. AVS is a common adopted basic standard in the important application of high definition digital television, network television, video communication and so on. In order to gain good coding result, AVS adopts so many new algorithms that the amount of calculation increase drastically. As a result, its applications in PC environment are restricted. In order to realize AVS real-time codec, many optimization works have to be done. But these optimization gains should not at the cost of obviously coding performance drop.
We mainly studied the AVS standards of video coding/decoding,used hardware and software resources of research team,adopted PC platform,designed a AVS video real-time decoding system.After post-optimization,the average resolution sequence is about 20.8 frame/s,basically achieved real-time decoding broadcast with little performance loss.Our work mainly focused on the following:
Firstly, we studied the AVS standards of video,including development situation,system structure,bit stream structure and its core technology.Through the analysis of the AVS standards,we also made a comparison between AVS standards and MPEG-4/H.264 technology.
Secondly,we finished the general design and implementation of an AVS real-time decoding system,mainly focusing on decoding system framework selection of software and hardware platform,system working mechanism and technology route.Through the analysis of key problems,we gradually finished several core modules like image head and straps decoding,macro block decoding,code streaming parsing and cross-frame forecast.
Third,post optimizing and overall system performance testing after decoding system programming.Optimization is performed in memory structure, C code and MMX.Aiming at the movement compensation and loop filtering modules that contain massivecomputation and complex , we use fast or predigest algorithm to cut down the computation amount and adjust the decoding function to enhance the efficiency.And we use MMX to optimize the modules that are frequently called but well formed ,which would bring a higher performance.Achieved an main profile of PC, real-time decoder decoder.
Finally,A comprehensive summary is given about this paper,and point out the direction of furthor research and improvement,and the promising application prospect of AVS.
引文
[1]石磊.AVS_M视频解码系统在嵌入式系统上的实现和应用研究[D].浙江:浙江大学,2006.
[2]闫冬.H.264/AVC编码器的优化设计及实现[D].成都:电子科技大学,2004.
[3]黄铁军.我国视频编码国家标准AVS与国际标准MPEG的比较.数字音视频编解码技术标准工作组.北京
[4]李春林.限失真视频编码技术的研究和算法优化[D].2004.
[5] ITU-T. Recommendation H.261: Video Codec for AudioVisual Services at p*64kbits, Dec.1990
[6] Draft ITU-T Recommendation H.263,Video Coding for Low Bitrate Communication, ITU-T,1997
[7] ITU-T and ISO/IEC JTC 1,Generic coding of moving pictures and associated audio information– Part 2:Video,ITU-T Rec.H.262– ISO/IEC 13818-2 (MPEG-2),Nov.1994
[8] MPEG-4 Video Group.Generic Coding of Audio-Visual Objects: Part2– Visual 14496-2,ISO/IEC JTC1/SC29/WG11 N2502a,FDIS,Atlantic City,Oct.1998
[9]ITU-T Rec.H.264/ISO/IEC 11496-10,Advanced Video Coding,Final Committee Draft,Document JVT-G050,March 2003,pp160-194
[10]张帆.H.264视频编码标准的研究与优化实现[D].北京:清华大学,2004.
[11]曾桂三.基于定点DSP的AVS编码算法实现及其优化[D].太原:太原理工大学,2008.
[12]魏昕,张承业.AVS和H.264的性能与编码对比[J].实用影音技术,2006,(9):49-53.
[13]数字音视频编解码技术标准工作组,http://www.avs.org.cn/, 2002-2006.
[14]李蕾.AVS解码器在C64平台上的优化及文字提取算法的研究[D].浙江:浙江大学,2007.
[15]毕厚杰.新一代视频压缩编码标准:H.264/AVC[M] .北京:人民邮电出版社,2005.
[16]高文,王强,马思伟.AVS数字音视频编解码标准[J].中兴通讯技术,2006,12(3):6—9.
[17]Cong Peng,Chao Huang,Ronggang Wang et a1.Architecture of AVS Hardware Decoding System[J].Hong Kong : Proceedings of 2004 International Symposium on Intelligent Multimedia,Video and Speeeh Processing,2004:306—309.
[18]王争,刘佩林.AVS帧内预测算法及其解码器的硬件实现[J].计算机工程与应用,2006,(19):80-83.
[19] GB/T20090.2-2006,先进音视频编码,第二部分:视频[S].AVS专家组信息技术,2006.
[20] LB.Yang,K.M.Yu,J.Li,S.P.Li,An effective variable block-size early termination algorithm for H.264 video coding,accepted by IEEE Trans.on Circuits and System for Video Technology.
[21]生滨.AVS视频编解码芯片关键部件的体系结构研究与设计[D].哈尔滨:哈尔滨工业大学,2006.
[22]柳金峰.基于DSP的AVS-M视频压缩算法的研究与实现[D].长春:吉林大学,2005.
[23] JIXiang-yang,ZHAO De-bin,GAO Wen,et a1.New Bi-prediction techniques for B pictures coding[C]//Proceedings of the 2004 IEEE International Conference on Multimedia and Expo(ICME72004),Vol 1.Jun.27-30,2004,Taibei,China.New York,NY,USA:IEEE,2004:101-104.
[24]田晓华.AVS视频编码中整数变换与熵编码研究[D].武汉:华中科技大学,2004.
[25]虞露,胡倩,易峰.AVS视频的技术特征.电视技术,2005,277(7):8.11.
[26]宋巍.基于TMS320DM6446的AVS编码器实现[D].太原:太原理工大学,2008.
[27] TI.TMS320DM6446 Digital Media System on-Chip.Literatuire Number:SPRS283, 2005.
[28]陈光法,姚立敏,虞露.AVS熵解码与dsp实现.电视技术,2004,10:43_46.
[29]洋利.AVS视频解码器在DM642上的实现及其优化方法的研究.长春:吉林大学,2009.
[30]王涛.AVS视频解码器C语言建模研究.天津:天津大学, 2008年5月.
[31]李晓瑜.实时AVS视频解码系统.成都:电子科技大学,2009年5月.
[32]刘东华.AVS-M移动视频标准技术研究.北京:中国科学院计算技术研究所.2006年10月.
[33]郑建安.AVS视频解码研究及其关键模块优化.厦门:厦门大学.2007年8月.
[34]丁志强,柳定一.MMX技术在视频压缩中的应用.《数字通信》,2000年第3期,P51-53,2002.
[35]穆翠英.利用MMX技术实现H.264实时解码器.太原:太原理工大学,2008年6月.
[36]张云,林钟.PC平台新技术MMX,东南大学出版社,1998.
[37]尚嫣雯,张刚.H.264滤波器的优化.太原理工大学学报2010年专辑.
[38]成运.实时H.264关键编码算法研究与实现[D].长沙:国防科学技术大学,2006.
[39]董志平,陈水仙,艾浩军.视频编码中环路滤波的SSE2优化方法研究.计算机工程与应用,2006年8月.
[40]姚志恒,李博扬,沈兰荪等.MMX技术在视频编码中的应用.计算机与网络,2001.24: 28-29.
[2]闫冬.H.264/AVC编码器的优化设计及实现[D].成都:电子科技大学,2004.
[3]黄铁军.我国视频编码国家标准AVS与国际标准MPEG的比较.数字音视频编解码技术标准工作组.北京
[4]李春林.限失真视频编码技术的研究和算法优化[D].2004.
[5] ITU-T. Recommendation H.261: Video Codec for AudioVisual Services at p*64kbits, Dec.1990
[6] Draft ITU-T Recommendation H.263,Video Coding for Low Bitrate Communication, ITU-T,1997
[7] ITU-T and ISO/IEC JTC 1,Generic coding of moving pictures and associated audio information– Part 2:Video,ITU-T Rec.H.262– ISO/IEC 13818-2 (MPEG-2),Nov.1994
[8] MPEG-4 Video Group.Generic Coding of Audio-Visual Objects: Part2– Visual 14496-2,ISO/IEC JTC1/SC29/WG11 N2502a,FDIS,Atlantic City,Oct.1998
[9]ITU-T Rec.H.264/ISO/IEC 11496-10,Advanced Video Coding,Final Committee Draft,Document JVT-G050,March 2003,pp160-194
[10]张帆.H.264视频编码标准的研究与优化实现[D].北京:清华大学,2004.
[11]曾桂三.基于定点DSP的AVS编码算法实现及其优化[D].太原:太原理工大学,2008.
[12]魏昕,张承业.AVS和H.264的性能与编码对比[J].实用影音技术,2006,(9):49-53.
[13]数字音视频编解码技术标准工作组,http://www.avs.org.cn/, 2002-2006.
[14]李蕾.AVS解码器在C64平台上的优化及文字提取算法的研究[D].浙江:浙江大学,2007.
[15]毕厚杰.新一代视频压缩编码标准:H.264/AVC[M] .北京:人民邮电出版社,2005.
[16]高文,王强,马思伟.AVS数字音视频编解码标准[J].中兴通讯技术,2006,12(3):6—9.
[17]Cong Peng,Chao Huang,Ronggang Wang et a1.Architecture of AVS Hardware Decoding System[J].Hong Kong : Proceedings of 2004 International Symposium on Intelligent Multimedia,Video and Speeeh Processing,2004:306—309.
[18]王争,刘佩林.AVS帧内预测算法及其解码器的硬件实现[J].计算机工程与应用,2006,(19):80-83.
[19] GB/T20090.2-2006,先进音视频编码,第二部分:视频[S].AVS专家组信息技术,2006.
[20] LB.Yang,K.M.Yu,J.Li,S.P.Li,An effective variable block-size early termination algorithm for H.264 video coding,accepted by IEEE Trans.on Circuits and System for Video Technology.
[21]生滨.AVS视频编解码芯片关键部件的体系结构研究与设计[D].哈尔滨:哈尔滨工业大学,2006.
[22]柳金峰.基于DSP的AVS-M视频压缩算法的研究与实现[D].长春:吉林大学,2005.
[23] JIXiang-yang,ZHAO De-bin,GAO Wen,et a1.New Bi-prediction techniques for B pictures coding[C]//Proceedings of the 2004 IEEE International Conference on Multimedia and Expo(ICME72004),Vol 1.Jun.27-30,2004,Taibei,China.New York,NY,USA:IEEE,2004:101-104.
[24]田晓华.AVS视频编码中整数变换与熵编码研究[D].武汉:华中科技大学,2004.
[25]虞露,胡倩,易峰.AVS视频的技术特征.电视技术,2005,277(7):8.11.
[26]宋巍.基于TMS320DM6446的AVS编码器实现[D].太原:太原理工大学,2008.
[27] TI.TMS320DM6446 Digital Media System on-Chip.Literatuire Number:SPRS283, 2005.
[28]陈光法,姚立敏,虞露.AVS熵解码与dsp实现.电视技术,2004,10:43_46.
[29]洋利.AVS视频解码器在DM642上的实现及其优化方法的研究.长春:吉林大学,2009.
[30]王涛.AVS视频解码器C语言建模研究.天津:天津大学, 2008年5月.
[31]李晓瑜.实时AVS视频解码系统.成都:电子科技大学,2009年5月.
[32]刘东华.AVS-M移动视频标准技术研究.北京:中国科学院计算技术研究所.2006年10月.
[33]郑建安.AVS视频解码研究及其关键模块优化.厦门:厦门大学.2007年8月.
[34]丁志强,柳定一.MMX技术在视频压缩中的应用.《数字通信》,2000年第3期,P51-53,2002.
[35]穆翠英.利用MMX技术实现H.264实时解码器.太原:太原理工大学,2008年6月.
[36]张云,林钟.PC平台新技术MMX,东南大学出版社,1998.
[37]尚嫣雯,张刚.H.264滤波器的优化.太原理工大学学报2010年专辑.
[38]成运.实时H.264关键编码算法研究与实现[D].长沙:国防科学技术大学,2006.
[39]董志平,陈水仙,艾浩军.视频编码中环路滤波的SSE2优化方法研究.计算机工程与应用,2006年8月.
[40]姚志恒,李博扬,沈兰荪等.MMX技术在视频编码中的应用.计算机与网络,2001.24: 28-29.